Robotic knee prosthesis to follow changes in real-time gait patterns

The knee joints of trans-femoral prostheses play an important role in compensating for the normal knee joint by controlling the swing phase. Even when the amputee changes the gait pattern, such as through changing speed or stride length, trans-femoral prostheses must perform a lower-leg swing that is similar to that of a normal person. Thus, in recent years, electronically controlled knee joints have been developed, and methods to dynamically reproduce the flexion angle of the knee joint have been proposed. However, since amputees wearing electronically controlled knee joints can only walk in accordance with the control parameters set in advance by the prosthetist, they cannot respond to changes in the gait pattern in real time. This leads to disrupted gaits. Therefore, in this study, in order to follow changes in gait patterns in real time and ensure a stable gait, we developed a robotic knee prosthesis. For this purpose, we proposed a walking model based on adaptive finite state machine that refers to the gait characteristics of the previous gait cycle. Finally, we conducted walking experiments with an amputee and showed the effectiveness of the proposed method by stabilizing the subject's gait.